Collision count in rugby union: A comparison of micro-technology and video analysis methods

Abstract

The aim of our study was to determine if there is a role for manipulation of g force thresholds acquired via micro-technology for accurately detecting collisions in rugby union. In total, 36 players were recruited from an elite Guinness Pro12 rugby union team. Player movement profiles and collisions were acquired via individual global positioning system (GPS) micro-technology units. Players were assigned to a sub-category of positions in order to determine positional collision demands. The coding of collisions by micro-technology at g force thresholds between 2 and 5.5 g (0.5 g increments) was compared with collision coding by an expert video analyst using Bland-Altman assessments. The most appropriate g force threshold (smallest mean difference compared with video analyst coding) was lower for all forwards positions (2.5 g) than for all backs positions (3.5 g). The Bland-Altman 95% limits of agreement indicated that there may be a substantial over- or underestimation of collisions coded via GPS micro-technology when using expert video analyst coding as the reference comparator. The manipulation of the g force thresholds applied to data acquired by GPS micro-technology units based on incremental thresholds of 0.5 g does not provide a reliable tool for the accurate coding of collisions in rugby union. Future research should aim to investigate smaller g force threshold increments and determine the events that cause coding of false positives.

Validity of a Microsensor-Based Algorithm for Detecting Scrum Events in Rugby Union

“…Microsensors and associated algorithms have been used to detect tackles in rugby league 14 with accuracy improving with greater impact forces and longer duration of events. 15 However, this technology has previously been shown to be less useful for detecting tackle events in rugby union 21 and Australian football 42 match-play. A possible explanation for the poor performance of the algorithm in Australian football and rugby union match-play is that the tackle algorithm was trained on rugby league players, to identify rugby league tackles.…”

“…Of the contact movements performed during regular match-play, scrum events occur around 25 times per game, while depending on playing position, each player will complete approximately 30 rucks and tackles per match. 5,11,[19][20][21] Given the need for more time-efficient and accurate methods of evaluating the incidence and physical demands of contact events in Rugby Union, this research sought to establish the validity of a microsensor-based algorithm for the automatic detection of scrum events during training and match-play. Based on the demonstrated capabilities of inertial devices to quantify other aspects of sports performance, 11,22 it was hypothesised that scrum events could be accurately detected using wearable microsensors.…”

Attacking 22 entries in rugby union: running demands and differences between successful and unsuccessful entries

“…Collisions were not included in this analysis due to current GPS technology's inability to validly quantify collisions in rugby union. 13 Reardon et al 14 have highlighted that current GPS technology under-and overestimates collision count, when compared to video analysis. Accelerations and decelerations were not included in our study, due to reported issues over the validity of quantifying these metrics with current GPS technology.…”

The relationship between collision metrics from micro‐sensor technology and video‐coded events in rugby union

“…These authors suggested that the higher g force used to detect collisions from backs compared with forwards was likely due to velocity into contact. 12 A more recently published study has demonstrated a high correlation (0.96) between MST coded and video-coded collision events in rugby league. 13 Player Load™ (Catapult Group International Limited, Melbourne, Australia), and its variants (Player Load™ Slow, Player Load™ 2D) have also been described in the literature as potential indicators of collision or contact loading in rugby codes.…”